THE PRIMA FACIE CASE IN
NON-JURY TRIALS
T
ROSCOE STEEFENt
mmn-poI in a law suit has always, and quite rightly, been a time
of testing. The plaintiff has presented his evidence, or so much as counsel has thought necessary, and the question is whether a sufficient showing has been made to require the defendant to go on with his case. Or, to put
the question in the usual way, has plaintiff made out a prima facie case? The
issue is raised in jury cases by a defense motion for a directed verdict;' in nonjury cases, by several devices, as by a motion for non-suit or, in Federal practice today, by a motion to dismiss "on the ground that upon the facts and the
law the plaintiff has shown no right to relief."'3 So much is clear enough, or
would seem to be. But what is not so clear is whether, in a non-jury case, the
trial judge should evaluate the plaintiff's evidence just as he would if the case
were being tried to a jury.
It may be that the idea of a prima facie case has no place in the non-jury
trial. The judge, there, is himself the trier of fact and, hence, or so the argument
runs, may as well proceed at once to "weigh" the plaintiff's evidence. 4 If it is
Tr
t John P. Wilson Professor of Law, University of Chicago Law School. The writer is happy
to acknowledge much help from talks with colleagues, and from student papers, especially
papers by Richard B. Berryman, John V. Gilhooly, Phillip Mullock and Stephen K. Waite.
I See, Hackett, Has a Trial Judge of a United States Court the Right to Direct a Verdict, 24
YAIE L.J. 127 (1914). As to the old practice of demurring to the evidence, see TAxER, PRELlthNAtY TREATISE ON EVlDENCE AT THE CommoN LAW 234 et seq. (1898). The procedure
was extensively re-examined by both the majority and minority in Galloway v. United States,
319 U.S. 372 (1942).
2See, Slocum v. New York Life Ins. Co. 228 U.S. 364, 394 (1912), a case dealing with the
power of a court to give judgment notwithstanding a jury verdict, where Mr. Justice Van
Devanter comments: "In what has been said we would not be understood as implying that a
motion for a compulsory nonsuit and a demurrer to the evidence axe equivalents of a request
for a directed verdict, for while they are sometimes spoken of as analogous to it, this only
means that for the purpose of each the evidence must be taken most strongly in favor of the
opposite party."
3FED. R. Civ. P. 41(b).
4The argument appears to have been first stated, and most fully, by Hoyt, C.J. in Lambuth v. Stetson & Post Mill Co., 14 Wash. 187, 190,44 Pac. 148,149 (1896). There, in sustaining a dismissal by the trial court at the end of plaintiff's case, Judge Hoyt said: "When the
trial is before a jury, the court cannot weigh the testimony upon a motion for a nonsuit, for
the reason that it cannot weigh it at any time; but, when the trial is without a jury, the
court must eventually weigh the testimony for the purpose of determining where the preponderance is, and there is no reason why it should not so weigh it at the earliest possible time,
when the rights of the plaintiff will not be cut off by its so doing; and when the plaintiff has
introduced all of his proof and rested, no right of his will be cut off if the court then determines
what has been proven. It cannot be presumed that plaintiff's case will be strengthened by the
evidence put in by the defendant."
1959]
THE PRIMA FACIE CASE IN NON-JURY TRIALS
found wanting, why should he not then make findings of fact and conclusions
of law, in the usual way, and forthwith direct that judgment be entered for defendant? Not only would that seem to be the expeditious way to do, but it is
said to be "just," for the plaintiff has had his day in court, and it cannot be
supposed that his cause would be helped, were the court to hear the defendant's
evidence.5 Moreover, even if the contrary were true, plaintiff has no vested
right to require defendant to put on his case.6 These were the arguments, sanctioned in several Circuit Court cases to be noticed later, which brought about
the 1946 amendment to Rule 41(b) of the Federal Rules.7 It is now made discretionary with a trial court, when sitting without a jury, either to deny the
motion, presumably upon the ground that a prima facie case exists, or to weigh
the evidence and dispose of the case at once.
Thus, the rule-makers chose to straddle the issue, and let the trial courts do
as they might think best. This, at least, was an easy solution-and not displeasing to the trial courts-though plainly it meant that any trial judge might,
at will, over-turn a practice of very long standing under which the rights of
plaintiff and defendant to a full hearing on the merits had been kept in fairly
even balance. Traditionally, courts have felt impelled to withhold judgment, in
non-jury as in jury cases, when a plaintiff has introduced "substantial" evidence
upon which a reasonable person might find in his favor.8 Perhaps there is in this
a certain wisdom born of long experience, for even plain cases sometimes take
on a different aspect when both sides are heard. 9 However that may be, the
5Porter v. Wilson, 39 Okla. 500, 135 Pac. 732 (1913). Plaintiff's claim here hinged on
whether his mother and father hadbeen married according to thecustoms of the Creek Indians.
It was urged that defendant's demurrer to the evidence admitted the facts tending to make out
plaintiff's case and that, if they supported the necessary allegations of the complaint, the
demurrer should be over-ruled. But the court held to the contrary; the trial judge, sitting
without a jury, was said to have acted properly in "weighing the plaintiff's testimony for the
purpose of determining the rights of therespective parties." Id. at 503. It is not clear whether
the trial judge thereby determined (1) that no reasonable person might find for the plaintiff
on the evidence when so "weighed," or (2) that the trial judge himself was not persuaded,
whatever other reasonable men might think. The point, if apprehended, was not considered
material, for the court said, "there being testimony reasonably tending to support the judgment of the court, the same will not be weighed by this court to ascertain whether the court's
decision is against the preponderance of the testimony." Id. at 508.
6Porter v. Wilson, 239 U.S. 170 (1915), is usually cited for this point. See Note, 45 MIcH.
L. REv. 788 (1947). Needless to say, Mr. Justice McKenna did not make so broad a statement;
the holding was that plaintiff's rights under the Fourteenth Amendment had not been shown
to have been violated by the action of the Oklahoma court.
7The third sentence of Rule 41(b) now reads: "In an action tried by the court without a
jury the court as trier of the facts may then determine them and render judgment against the
plaintiff or may decline to render any judgment until the close of all the evidence."
" See: M1aryland Casualty Co. v. Jones, 279 U.S. 792 (1929); Jefferson Electric Mfg. Co.
v. United States, 291 U.S. 386 (1934); Gunning v. Cooley, 281 U.S. 90 (1930).
9 See comment of Justice Van Orsdell in Catholic University of America v. Waggaman, 32
App. D.C. 307, 320 (1909): "While we recognize therule that it is incumbent upon the plaintiff
to establish a prima facie case by affirmative evidence before the defendant can be put upon
THE UNIVERSITY OF CHICAGO LAW REVIEW
[Vol. 27
amended rule encouraged, if it did not require, trial judges to reach a decision
at once. And the result, it would seem, has been increasingly to dismiss without hearing the other side; but that this should be so is perhaps the purpose of the new procedure.
Such in brief are the arguments for change. But surely so venerable an institution as the prima facie case in non-jury trials is not to be dispatched that
summarily. At most the proponents of the new procedure have made a prima
facie case; and, under the traditional practice, the other side is entitled to be
heard. At all events, it is proposed in what follows to look first in more detail
at how and why the new rule came about, and then to examine into some of
its implications, which, if considered, do not appear to have been given sufficient
weight by the rule-makers. In other words, it is proposed to see what sort of
a case can be made for the prima facie rule."0
CONSTRUCTION OF RULE
41(b)
The first court to construe Rule 41(b), prior to the 1946 amendment, had
little doubt as to its meaning. The case, FederalDeposit Ins. Corp. v. Mason,"
was brought to recover the amount of certain dividends alleged to have been
unlawfully declared and paid out by the defendants. After reviewing plaintiff's
evidence, judge Maris held that the lower court had erred in dismissing the
case: "Inasmuch as an order of dismissal under Rule 41(b) is the equivalent of
a directed verdict we must upon review of such an order view the evidence and
all inferences reasonably to be drawn therefrom in the light most favorable to
the plaintiff."' 2 And, in that light, plaintiff was said to have made out a case
sufficient to go to a jury, hence, sufficient to require the trial court, sitting
without a jury, to hear defendant's evidence before reaching a decision.
There seems to be little doubt that the rule-makers regarded the motion to
dismiss and the motion for a directed verdict to be controlled by the same procedure. Indeed, the Advisory Committee, in its note to Rule 41(b), said that
"it provides the equivalent of the directed verdict practice for jury actions
which is regulated by Rule 50."" And, at the New York Symposium, when
asked whether a court must make findings of fact, either upon granting a
motion for dismissal or for a directed verdict, Mr. William D. Mitchell was
his proof, yet, where the plaintiff has established a state of facts from which the jury can draw
reasonable inferences tending to sustain the contention of the plaintiff, it is error for the court
to dispense with the services of the jury. Especially is this true where the evidence of the defendant must, of necessity, be such as to illuminate the issues and assist the jury in arriving at
a proper and just verdict." But then, things moved at a slower pace in 1909.
10The writer's initial bias in this matter came when he argued and lost the point in United
States v. United States Gypsum Co., 67 F. Supp. 397, 417 (D.D.C. 1946).
1 115 F.2d 548 (3d Cir. 1940).
2Id. at 551.
1328 U.S.C.A. 204 (1958).
1o99
TEE PRIMA FACIE CASE IN NON-JURY TRIALS
quite explicit: "The answer is 'No'. Rule 52 with regard to findings is based
on the assumption that you will pay a little regard to the practice in the courts
for the last 150 years.' 4 Mr. Mitchell went on to say that a case dismissed upon
motion was not one "tried upon the facts,"u S and so, literally speaking, would
not come within Rule 52. No one supposed, it seems, that a trial court would
ever attempt to "weigh" plaintiff's evidence, when passing on such a motion.
But in Gary Theatre Co. v. Columbia PicturesCorp.,'6 an anti-trust case, the
trial court, sitting without a jury, purported to have done that very thing.
That is to say, at the end of plaintiff's case the court proceeded to weigh the
evidence and make findings of fact and conclusions of law. Moreover, judge
Lindley, on appeal, saw nothing strange in this. Without regard to prior practice; without reference to the probable purpose of the rule-makers; he refused
to consider, as he was urged to do, whether plaintiff had made out a prima facie
case: "Under Rule 41(b) the judgment, supported by findings, was an adjudication upon the merits, inasmuch as defendants moved for dismissal upon the
ground that, upon the facts and the law, plaintiff had shown no right to relief."' 7
Hence it followed that, under Rule 52(a), the judgment could not be reversed,
unless the trial court's findings of fact could be set aside as "clearly erroneous."
And that, "we do not believe we are justified in doing."' 8
The case of Young v. United States, 9 decided a year earlier in the Ninth Circuit, is usually cited as supporting the same conclusion. But, there, the defendant had rested his case at the end of plaintiff's evidence. In such circumstance, obviously, the lower court's judgment and findings of fact were upon
the whole case; the time to inquire whether plaintiff had made out a prima facie
case had passed. Nonetheless, Judge Wilbur appears to have reasoned the case
as one calling for a construction of Rules 41(b) and 52(a). "These rules," he
said, "do away with the former distinction," in California practice, "between a
judgment of nonsuit and a judgment on the merits in all suits or actions where
the court sits without a jury. ' 20 It followed that the case could be reversed only
if the trial court's findings were "clearly erroneous." The lower court was
affirmed.
These cases, it will be noted, did not pass on the question as a matter of
policy; they merely held that a trial court had power to weigh plaintiff's evi'4PROCEEDINGS,
(1938).
15Id.
INSTITUTE ON THE FEDERAL
RULEs, Washington and New York, 336-37
16120 F.2d 891 (7th Cir. 1941).
7
Id.at 892. The court may have been confused by the thought that, because the dismissal
was "on the merits," findings were required. Rule 41(b) provides, however, that any dismissal
(except for lack of jurisdiction or improper venue) "operates as an adjudication upon the
merits."
18 .1d.at 894-95.
19111 F.2d 823 (9th Cir. 1940).
20Id. at 825.
THE UNIVERSITY OF CHICAGO LAW REVIEW
[Vol. 27
dence and make findings of fact when ruling on a motion to dismiss. But in
Bach v. Friden CalculatingMack. Co., 21 a Sixth Circuit case decided in 1945,
Judge Hamilton went further. Asserting, first, that "the purpose" of the rules,
as if there were only one, "is to expedite the trial of cases," he went on to say
that the question was one of "fact," and that the "sensible course" for a trial
court is to dismiss if plaintiff has not made out his case "by a preponderance of
evidence."2 Nor would appellants be hurt if that were done;
they have advanced their whole cause by putting in all of their evidence and it would
be a refinement of technicality to say that such evidence and all reasonable inferences
to be drawn therefrom must be viewed in the light most favorable to the appellants
and if -there be any substantial evidence supporting their claim, the court must put
3
2
the appellees to proof.
Again, the lower court was affirmed.
Then in 1946, Judge Stephens, sitting as a trial judge in United States v.
United States Gypsum Co.,2 4 put the matter even more strongly. Not only would
it be proper for a trial court to weigh plaintiff's evidence under Rule 41(b), but,
he said, "it is the duty of the court to weigh the evidence, to draw inferences
therefrom and, if it finds the evidence insufficient to make out a case for the
plaintiff, to render a decision for the defendant on the merits. '2 5 Moreover,
Judge Stephens thought the weighing should be done with care; he took more
than two years in which to do the job, wrote a scholarly 120 page opinion, made
some 118 detailed findings of fact, and stated a dozen conclusions of law2 It is
not clear whether the rule-makers relied on this decision also, when amending
Rule 41(b), but it gave unqualified support for their action. Once again, the
plaintiff's case was dismissed.
Thus, when the question reached the rules committee, the clear "weight of
authority," at least, favored a construction which permitted a trial court to
weigh the plaintiff's evidence. And, likewise, to make findings of fact and conclusions of law. The amendment to Rule 41(b) made it clear that this was a
proper procedure. But it also provided quite as clearly that a trial court, if it
thought best, might "decline to render any judgment until the close of all the
evidence." 27 That is, if a trial court should determine that there was substantial
evidence in the record upon which a reasonable person might decide for the
21148 F.2d 407 (6th Cir. 1945).
23Id. at 411. (Emphasis added.)
24 67 F. Supp. 397 (D.D.C. 1946).
at 410-11.
at 419-420. (Emphasis added.) judge Stephens noticed thepoint that therule-makers
had said the motion to dismiss was the equivalent of a motion for a directed verdict, but, he
said, this was true only in the sense that, in each case, if the motion was denied the defendant
might proceed to put in his case. It was therefore said to have no bearing on the issue before
the court. Any other construction, he said, would "ignore" the fact that in non-jury cases the
judge, himself, is the "trier of the facts."
26
This experience alonewould suggest that not much court time is saved by the "weighing"
procedure.
" FED. R. Crv. P. 41(b).
22Id.
2Id.
1959]
THE PRIMA FACIE CASE IN NON-JURY TRIALS
plaintiff-whether or not the court had reached that decision for itself-the
defendant's motion to dismiss could properly be denied.
Nothing was said, at least not directly, upon whether a trial court (without
weighing the evidence) might still dismiss for the reason that plaintiff had
failed to make out a prima facie case. That procedure was simply given a cold
shoulder, possibly to encourage trial courts increasingly to make findings of
fact, and so to facilitate appellate review. But it must be remembered that the
motion to dismiss under Rule 41(b) proceeds on a dual ground, that is, "upon
the facts and the law." Thus, clearly, a trial court may properly dismiss a case
solely as a matter of law, 28 and a disposition on that ground operates "as an
adjudication upon the merits" within the rule, quite as much as if the court had
dismissed the case as a "trier of the facts," after weighing the evidence. Therefore, since the question -whether plaintiff has made out a prima facie casehas always been held to be one of "law, '29 there is no real doubt that a trial
court may still properly grant a motion to dismiss on that basis, without making findings of fact. The case then is not in any real sense "tried" on the facts.
EXPERENTAL TRIAL DATA
It has been blandly assumed up to this point, as is customary with lawyers,
that more cases would be dismissed under the "weighing" test, than under the
"substantial" evidence rule. Indeed, the proponents of the "weighing" test
regard this not only as fact, but as a virtue. Perhaps they are wrong on their
basic assumption; certainly a person who has not experienced the stress and
anxiety of presenting a closely contested case in court may well regard the
whole thing as a matter of words, and say that no significant difference would
occur. Who knows how judges actually dispose of cases, they ask, whatever
formula is used? This is a healthy attitude, I suppose, but it leads exactly
nowhere.
At all events, in order to get some quantitative idea of what, if any, difference
might be expected, the following experiment was set up. 0 The sound track of a
mock presentation of an actual negligence case, replete with question, answer,
28 For example, plaintiff may have failed to state a cause of action. Or, as in Oscanyan v.
Arms Co., 103 U.S. 261 (1880), plaintiff may have disclosed in his opening statement that he
had no lawful right to recover. Findings in such cases would be superfluous, if not impossible.
29See Maryland Casualty Co. v. Jones, 279 U.S. 792, 795 (1929), ajury-waived case, where
Mr. Justice Sanford said: "The motion for nonsuit-which corresponded to a motion for a
directed verdict-presented the question whether the evidence, with every inference of fact
that might be drawn from it in favor of the plaintiff, was sufficient in matter of law to sustain
a judgment. See Central Transp. Co. v. Pullman's Car Co., 139 U.S. 24, 38. This presented a
question of law which is reviewable. ..."
3'A team of lawyers and social scientists at the University of Chicago Law School is currently engaged in a major study of the American jury system, under a grant from the Ford
Foundation. One principal technique of investigation has been the experimental jury. With
the cooperation of the courts in several cities, recorded mock trials- based on the transcripts
of actual cases-have been submitted to juries drawn from the regular jury panel and have
been decided by them in the atmosphere of the courtroom. It has thus been possible to utilize
the regular adult jury population, to try the same case repeatedly, to test experimental varia-
THE UNIVERSITY OF CHICAGO LAW REVIEW
[Vol. Z7
objection and ruling, arguments of counsel, the trial court's instruction, and so
on, was played to a college class of 48. Each member of the class, acting individually, was asked to reach a decision upon the merits at the end of the case.
But, midway, a motion to dismiss was made, and each member was instructed
to pass upon it in the role of a trial judge. The instructions to 30 members of
the class were to determine, individually, whether there was substantial evidence so far presented upon which a reasonable person might decide in favor of
the plaintiff. The members of the other group, 18 in number, 3' were separately
instructed to go a step further, that is, to weigh the evidence in accordance
with the court's instructions and to decide, each for himself, whether plaintiff
in fact had sustained the burden of proof and made out a case on the merits.
It is not suggested, of course, that such an experiment necessarily "proved"
much of anything. For, however careful the experimenter may be to follow
actual court procedure, it still is true that one is an experimental trial, the other
an actual one. Nonetheless, the results did confirm the trial lawyer's hunch,
that there is a very real difference between the two tests. Of the 30 students
who used the "substantial" evidence test, 25 were for denying the motion to
dismiss, while only 5 granted it. On the other hand, of the 18 who were instructed to "weigh" the evidence, 10 were for denial, while 8 granted the
motion. For the benefit of the lay reader, the ratio in the first case, thus, was
1 to 5 for dismissal, in the second, 4 to 5. That would seem to be quite a variance. 3' Stated percentage-wise, 17 per cent in the first group were for dismissing
the action, and 44 per cent, in the second. For those readers who gauge the
soundness of their judgment by the conventions of statistical inference, the difference, of 27 per cent, is significant at the .05 level.33
tions, and to record in full the deliberations of the jury on the experimental case. The experimental jury work has been conducted by Professor FredL. Strodtbeck of the Social Psychology
and Sociology faculties of the University of Chicago in collaboration with Professor Harry
Kalven, Jr., of the Law faculty.
The experiment described in the text was administrated by Professor Strodtbeck and
represents an adaptation of one of the experimental scripts-the so-called Conway case, dealing with the possible liability of a manufacturer and a retailer for injuries caused by a defective vaporizer. The experimental jury researches are now in the process of write-up for publication in book form. A full report on the Conway case will be included therein.
31The difference in numbers between the two groups was a matter of chance.
32Later the same experiment was tried with a law class which had had two quarters of
schooling. The direction of the differences was the same as for the college class, but the extent
was less, and not enough to be statisticaLly significant. Perhaps this could be explained by
the saying, a little law is a dangerous thing. But it might well be accounted for by one change
in the format; the members of the law class did not receive their instructions, to "weigh" or
not, until midway; the members of the college class, on the other hand, received their instructions at the start, and so, no doubt, were busily "weighing" or testing for "substantiality"
as they listened to the evidence. This, of course, conformed more closely to trial practice, as
it may be assumed the trial judge would know at the outset which rule he intended to apply.
3"A significance level of .05 or .01 represents the risk that one takes that one's decision is
wrong. The arithmetical operations for testing a statistical hypothesis of the type here used
are described in all modern introductory statistics texts in connection with the chi-square dis-
19s9]
THE PRIMA FACIE CASE IN NON-JURY TRIALS
But does it follow that there would be any important difference in ultimate
result? Perhaps the appellate machinery would return those cases, if any, which
were improperly dismissed, so that, except for the delay and expense involved,
plaintiff would not have been hurt. Before taking up this question, let us first
see what happened when the dismissed cases were carried on to a conclusion.
Of course, in actual practice, data of this sort is not available. But on an experimental basis it was a perfectly proper thing to ask each student to give his
decision upon the whole case at the end, regardless of how he may have ruled
on the motion to dismiss.
The results, here too, were interesting. Of the first group, that is, the one
using the "substantial" evidence test, only 1 of the 5 who had voted to dismiss
changed his mind and held for the plaintiff. But of the second group, 3 of the 8
who had dismissed after "weighing" the evidence changed their views, and voted for the plaintiff. 34 That is, not only was a larger percentage of cases dismissed
under the "weighing" test, but a larger percentage of the cases so dismissed
were later decided for plaintiff, after all the evidence was in for both sides. But,
the difference in result between the two tests was only 18 per cent, and, we are
3
reminded, that is not quite a statistically significant difference. 1
tribution, see for example, DIXON & MASSEY, INTRODUCTION TO STATISTICAL ANALYSIS, 18491 (1951). The line of inference is first to assume that the two sets of students have been
drawn from a population in which the percentage who "deny" is the same as that of the two
sets combined. Then, by reference to standard distributions, it is possible to determine how
frequently a difference of the observed magnitude would have been expected to occur. This
assumes repeated samples of the size here used from a very large population. If the difference
in question would have occurred less than one time in twenty trials, then one rejects the null
hypothesis that the samples were drawn from the same population. Proceeding on this basis,
one assumes that the samples were from different populations which had different dispositions
to "deny." This assumption is right 95 per cent of the time and wrong 5 per cent of the time.
If the criterion for rejecting the null hypothesis were more stringent, thereby reducing the 5
per cent, one would increase the frequency of errors of a second type: namely, the error of
saying that two samples were from the same population, when, in fact, they were from different populations. It should be clear that, useful as they are, significance tests, and other
related statistical criteria do not go to the problem of generalizing from what students do to
the behavior of judges. (Per, Prof. Strodtbeck.)
34 We started with an hypothesis that, if a judge were to actually make up his mind to
decide for the defendant-on the basis of plaintiff's evidence alone-it would be more difficult for him, should he be reversed on appeal, to hear the defendant's evidence fairly, than if
he had merely examined plaintiff's evidence in the first place under the "substantial evidence"
test. But this was not borne out, though not disproved either. See generally, FEsTINGFR, A
THEoRY or CoorIqi~v DISSONANCE (1957).
15Perhaps, though, it is legally significant. See Judge Frank, dissenting in United States v.
Johnson, 238 F.2d 565, 571-572 (2d Cir. 1956), a case where the majority had denied a motion
to appeal informapauperis:"The way out of this apparent dilemma is to consult the interest
of justice: Surely, even if but one out of a hundred attempted appeals by indigents has merit,
justice compels the conclusion that that appeal shall be heard. It is no answer that so many
appeals ill result as to 'crowd the docket.' If so, more judges should be appointed. True, the
cost of running the government will somewhat increase. But I, for one, cannot sleep well if I
think that, due to any judicial decisions in which I join, innocent destitute men may be behind
bars solely because it will cost the government something to have their appeals considered."
THE UNIVERSITY OF CHICAGO LAW REVIEW
[Vol. 27
What, then, of appellate review? That is a matter to be examined later, in
the light of the case law. But in the cases dismissed upon the "substantial" evidence rule, it may be assumed the appellate courts would use the same test in
their deliberations. 3 It is the situation where the trial court has "weighed"
plaintiff's evidence, and found that he failed to sustain the burden of proof,
which gives concern. One's hunch, here, is that a much smaller number of
reversals would occur 7 But perhaps that is as it should be; at any rate, that
is the question.
It should be noted that this experiment dealt with a closely balanced case. 33
But it is in just such cases, it would seem, that the judicial machinery should
be most carefully adjusted. In the fringe cases, where a plaintiff has either failed
to put in any proof at all or, on the other hand, has presented over-whelming
evidence, it should not make too much difference which rule is used by the trial
court. But plainly it will not do to argue from such cases-as some proponents
of the "weighing" rule appear to do-that, in every case, when a trial judge has
made up his mind on the merits adversely to plaintiff, he should grant defendant's motion to dismiss the case and direct a judgment. Or, that the appellate
machinery should be geared to sustain his judgment.
A REn
.r oF TECHrNICALITY
It would be well now to look more carefully at the role of the trial judge
when sitting without a jury. The experimental data just discussed indicate that
many more actions will be dismissed under the "weighing" test than has been
true under the "substantial" evidence rule. And the data suggest, at least, that
some of these will be dismissed unfairly. How, then, is one to explain the
change? It surely is no answer to point out-with the air of a man who has
just discovered fire for the first time--that in the non-jury case the judge himself is a trier of fact. It is not complimentary to prior courts to suggest that they
really were not aware of their dual role when trying a case without a jury. Nor
is it any more so to say that they probably were misled "by the unnoticed process of generalization"; 39 that is, since the "substantial" evidence rule was used
when passing on a motion for a directed verdict, they just supposed the same
rule must apply in all cases.
The best affirmative argument, it seems, is the presumed artificiality of
16Federal Deposit Ins. Corp. v. Mason, 115 F.2d 548 (3d Cir. 1940).
37 It is one thing to test the trial court's action by asking whether there is any substantial
evidence (which a jury or an appellate court might believe) to support a judgment for plaintiff,
and a very different one to ask if there is substantial evidence to support the court's action in
dismissing the case. One tends to hold judgment in abeyance until the evidence is all heard,
the other to close the case out as soon as may be.
"8Of the 50 juries who heard the case, in the course of the experimental jury study, 22
found for the plaintiff against the manufacturer alone, 1 against the distributor alone, and 10
against both defendants.
39 Note, 45 McH. L. REv. 788, 789 (1947).
19591
THE PRIMA FACIE CASE IN NON-JURY TRIALS
applying the "substantial" evidence test,4" in non-jury trials. The demand of
Rule 1 for a "speedy" determination of every action is surely a subsidiary
one.41 At all events, Judge Stephens puts the argument on artificiality as well
as any one. "It is not reasonable," he said, "to require a judge, on motion to
dismiss under Rule 41(b), to determine merely whether there is a prima fade
case," for that would compel "a defendant to put on his case and the court to
spend the time and incur the public expense of hearing it... even though the
judge had concluded that on the whole of the plaintiff's evidence the plaintiff
ought not to prevail."42
So stated, the argument has a certain plausibility. There are, of course, many
cases where a plaintiff may have wholly failed to establish his case, or an essential part of it, which comes to the same thing. Defendant's motion then should
be quickly granted. But why would it not suffice to dismiss upon the the ground
that plaintiff had failed to put in "substantial" evidence to support a judgment
in his favor? In many cases, no doubt, it would. But in the closely balanced case,
it seems, it would not. There the trial judge may be quite sure, in fact, that
plaintiff las put in "substantial" evidence to support his case; that other reasonable men, to-wit: a jury or an appellate court, might well decide in his favor;
but the trouble is that the judge himself has "concluded" that "plaintiff ought
not to prevail." 4 3 Hence, he decides to "weigh" plaintiff's evidence.
This puts a somewhat different face on the matter. Perhaps the trial court
holds strong views opposed to the policy of the existing law. Perhaps, the trial
court's new-found interest in his role as a trier of fact is a move to give his determinations, half way through a case, that degree of finality which is accorded
a jury verdict at the end. And, by so much, to make it the more difficult for
an appellate court to reverse. This, of course, cannot be true, even though no
trial judge likes to be reversed.
But, still, why are trial courts so anxious to "weigh" a plaintiff's evidence?
There is no gainsaying the point that the weighing test is highly advantageous
to a defendant. It gives him two bites at the apple, so to speak, where heretofore he has had but one. At the end of plaintiff's case he may say: "I know that
plaintiff has put in 'substantial' evidence, but perhaps on a 'weighing' the trial
court will dismiss, in which case, at the very least, I will have gained one or
two years delay. If the case is returned, I will go ahead and put in my evidence,
41The writer is quite aware that there are different views as to the scope of this test, varying, it seems, on one's regard for trial by jury. See the majority and minority opinions in
Galloway v. United States, 319 U.S. 372 (1943), and criticism by Prof. McBaine, Trial.Practice, 31 CALIF. L. R.v. 454 (1943). For present purposes it is not necessary to determine the
point, for, whatever its scope, the test is plainly distinguishable from a "weighing" test.
4
1FD
R. CIrv. P. 1, provides three tests: "They [the rules] shall be construed to secure the
just, speedy and inexpensive determination of every action."
4
2United States v. United States Gypsum Co., 67 F. Supp. 397, 418 (1946).
4
3Id. at 418.
THE UNIVERSITY OF CHICAGO LAW REVIEW
[Vol. 27
since I have not been required to rest. Then, at the conclusion, I can again have
the evidence 'weighed.' Clearly there is everything to gain, and nothing to lose."
One hesitates to raise the question, for someone may think it a fine idea, but
why should we not adopt a mid-way "weighing" procedure in jury cases? The
jury, too, is a trier of facts, and if it could be persuaded to "conclude," after
hearing only the plaintiff's evidence, that he had failed to sustain the burden
of proof, why bother the defendant with going on with his case? Why spend
more of the taxpayer's money, when-if the argument is sound-it cannot be
supposed that plaintiff would be helped by defendant's evidence, or that a
juror, any more than a judge, could ever see the case in better perspective on
hearing both sides?
The answer, probably, lies in a deep-seated feeling as to what is fair. Of course,
defendant may not be required to put on his evidence, but since he ordinarily
will have evidence which may throw light on the points at issue, the jury has
always been expected to refrain from concluding anything until his case is in.
Why, then, should a judge, as a trier of fact, be encouraged to act with any less
impartiality-unless, indeed, the need to hurry and clear his docket is of paramount importance?
The Court of Appeal in Alexander v. Rayson" was quite clear how the question should be answered: "Certainly no one would ever dream of asking a jury
at the end of plaintiff's case to say what verdict they would be prepared to give
if the defendant called no evidence, and we fail to see why a judge should be
asked such a question in cases where he and not a jury is the judge that has to
determine the facts.
45
The trial judge there had been asked, not to "weigh" the evidence, but merely
to rule that, "accepting the evidence given on the defendant's behalf, there was
no case to answer in law."' 46 It was decided that the trial court had discretion to
grant such a motion, without requiring the moving party to rest. But the court
said this was not only an irregular but a most inconvenient procedure: "For the
judge in such cases is also the judge of fact, and we cannot think it right that
to express any opinion upon the evidence until
the judge of fact should be asked
47
the evidence is completed.
"WEIGn1NG" PLAINTF.F'S EVIDENCE
It has always been very mysterious to me how a judge goes about "weighing"
a plaintiff's evidence. What does he weigh it against? Perhaps one's thinking is
shaped by the picture of a scales of justice, with plaintiff's evidence on one side,
44 [1936] 1 K.B. 169.
45Id. at 178.
46
Id. at 178. This motion, it will be noted, asks for a ruling according to the "substantial"
evidence test. It was made to test defendant's proof on an affirmative defense. No English
court appears yet to have had the temerity to "weigh" evidence midway in a trial.
47
d. at 178.
1959]
THE PRIMA FACIE CASE IN NON-JURY TRIALS
and nothing much on the other, that is, unless the assertions of defense counsel
and the court's own doubts are to be put there. Of course, where a defendant
has rested his case, the court perforce must reach a decision. But two things are
noteworthy then. In the first place a defendant will ordinarily not rest unless
he is quite sure of his ground.48 Thus, as a practical matter, a court will not
often be asked to decide a close case at mid-point. In the second, the very failure
of the defendant to call any evidence, when his version of the facts might be
helpful to the court, is a tacit invitation to view the case in a light favorable to
49
plaintiff.
So also, it is difficult to see how it is meaningful to speak of the burden of
proof in these cases. Of course the onus is upon plaintiff to put in substantial
evidence to support the necessary allegations of his complaint," but that is all;
there is no question of burden of proof in any other sense, for surely even slight
evidence preponderates over none at all."' However, the courts which purport to
"weigh" the plaintiff's evidence carry through, and insist that plaintiff must
somehow sustain a burden of proof. Judge Hamilton, it will be recalled, said in
the Bach case that unless the plaintiff had made out his case by "a preponderance of evidence" 2 the motion to dismiss should be granted. He did not explain
how the thing was to be done.
In United States v. Morgan,53 Judge Medina not only decided that he must
weigh the plaintiff's evidence to determine whether the burden of proof had
been met, but took the position that the evidence should be weighed as it was
introduced:
Don't forget that the pieces of the mosaic are constantly forming. The talk about
putting the mosaic together at the end of the case means putting true, accurate,
factual pieces together after the controversy on the facts has been resolved as to each
48 If defendant has no case at all, and is merely bluffing, the action of a court in strictly
"weighing" plaintiff's evidence becomes a travesty, except to someone immersed in the old
view that a law-suit is a sort of game.
49 Kirby v. Tallmadge, 160 U.S. 379 (1895). The defendants here, in an action to remove a
cloud upon title, had rested without introducing any evidence, and, when the decree went
against them, brought this appeal. In affirming the lower court, Mr. Justice Brown said: "As
they had it in their power to explain the suspicious circumstances connected with the transaction, we regard their failure to do so as a proper subject of comment. 'All evidence,' said
Lord Mansfield in Blatch v. Archer, Cowper 63, 65 (1774), is to be weighed according to the
proof which it was in the power of one side to have produced and in the power of the other
side to have contradicted.'" Id. at 383. See also, The Eastchester, 20 F.2d 357, 358 (1927).
10See, 9 WIGmoix, EvmENcz § 2494 (3rd ed. 1940).
51In Quock Ting v. United States, 140 U.S. 417 (1890). Mr. Justice Field found "inherent
improbability" in the testimony of petitioner (a Chinese boy held for deportation), so it could
be disregarded, but he stated the general rule to be that "positive testimony as to a particular
fact, uncontradicted by any one, should control the decision of the court.... " Id. at 420.
12 Bach v. Friden Calculating Mach. Co., 148 F.2d 407, 410 (6th Cir. 1945). See text accompanying note 22 supra.
53 118 F. Supp. 621 (E.D.N.Y. 1953). The charge here was a conspiracy to fix non-competi-
tive prices in the purchase and sale of securities.
THE UNIVERSITY OF CHICAGO LAW REVIEW
[Vol 27
piece. So you must not think that I am going to wait until perhaps a year from now to
decide out of a huge welter of proof what are the pieces to be put together,.., as to
a substantial number if I find the little piece to be black instead of white, how is it
going to fit in.M
As a matter of conspiracy law, of course, this was plain error. For, as Mr.
Justice Van Devanter pointed out in United States v. Patten:55 "It hardly needs
statement that the character and effect of a conspiracy is not to be judged by
dismembering it and viewing its separate parts, but only by looking at it as a
whole." 56 And while the Justice's statement was made with respect to the sufficiency of a pleading, it would seem to apply, a fortiori, to a piecemeal disposition of plaintiff's evidence, whether disposed of for lack of "weight," or for a
supposed lack of "credibility." As a matter of evidence, though, how could the
trial court resolve "the controversy on the facts" before the other side had been
heard, unless, indeed, he gave weight to the unsworn statements of defense
57
counsel?
In the Gypsum case,5 8 judge Stephens also held that the burden of proof was
on the plaintiff. He then went on to elaborate certain "principles of proof,"59
which he said applied, though they were all derived from cases where both sides
had been fully heard. Thus the "equivocal" evidence test was taken from
5
Pennsylvania R.R. Co. v. Chamberlain."
The court, there, in determining
whether to grant a motion for judgment notwithstanding the verdict, said this:
We, therefore, have a case belonging to that class of cases where proven facts give
equal support to each of two inconsistent inferences; in which event, neither of them
being established, judgment, as a matter of law, must go against the party upon whom
rests the necessity of sustaining one of these inferences as against the other, before
he is entitled to recover. 1
Such a rule is perhaps too hypothetical to be of much use, even at the end
of a trial, except as a mischief maker.12 But, however that may be, there would
seem to be no proper basis for using it in the middle of a non-jury case, especially where the case is a close one. That is pre-eminently not a time when the
trial judge should "conclude" anything, one way or the other. For it may be
5
4Id.at Record, pp. 8518-19. See Steffen, Tle Investment Bankers' Case: Some Observations,
64 YAx LJ.169 (1954); Whitney, The Investment Bankers' Case-Includinga Reply to Professor Steffen, 64 Yale L.J. 319 (1955).
- 226 U.S. 525 (1913).
6 Id. at 544.
57 The defendants were permitted to take more than three months in their opening statements, at the start of the case, to comment at length upon the government's exhibits, the
"true" nature of the business, and so on and on. It is not humanly possible that some of these
arguments did not get put in the scales to be "weighed" against the plaintiff's proof.
5
8 United States v. United States Gypsum Co., 67 F. Supp. 397 (D.D.C. 1946).
59
Id.at 449.
60288 U.S. 333 (1933).
1Id.at 339.
2
6 See Comments of Mr. justice Black, dissenting in Galloway v. United States, 319 U.S.
372, 405 (1943). Whether two possible inferences are so evenly balanced as to be precisely
equal is surely a matter for the jury to determine.
1959]
TBE PRIMA FACIE CASE IN NON-JURY TRIALS
assumed, if the defendant were to put on his evidence, the matter would
quickly be got off dead center, perhaps in defendant's favor, but by no means
necessarily so. The fair thing, therefore, would seem to be to put the defendant
to his proof, that is, unless he would prefer to rest his case.
6
The court's use of the "equivocal" evidence test in the Gypsum case
3
will
illustrate the point. The charge there, in general terms, was that defendant
manufacturers had conspired to organize their industry and stabilize prices by
the use of common patent license agreements. Certain exhibits, taken from the
files of National Gypsum, were introduced as tending to show the alleged plan,
and that company's connection with it. But, as to these, Judge Stephens said:
The recital in the minutes of the May 14, 1929, directors' meeting of National of the
chairman's statement that he had been "informed that all other manufacturers of
gypsum products... except... American had agreed to sign a license contract... ,"
and the resolution authorizing execution of license agreement with USG in the form
submitted.., are equivocal. It is without dispute that the various companies were at
this time considering the May 1929 licenses. It is no more inferable from these minutes
that the various (prospective) licensees had agreed with each other to sign the license
64
contract with USG than it is that they had agreed severally with USG to do so.
This example alone should show that a "weighing" procedure is itself quite
artificial, when only one side has been heard. Indeed, it is hard to imagine any
fact from which conflicting inferences may not be drawn. If the trial court is
unable to imagine any, defense counsel are only too ready to supply some
plausible ones.5 To illustrate, the evidence in the Gypsum case showed that all
but one of defendant manufacturers had met together in Chicago when the last
license agreements were signed and, more important, that USG had issued its
first general price bulletin at that time. One robust inference from this was
that defendants had thereupon reached a common agreement to organize their
industry and stabilize prices, but the court pointed out-at the suggestion of
defense counsel-that since there was a "most favored nation" clause in the
agreements, another inference was that they may merely have met together to
see that one licensee was not getting better terms than another. Whatever one
may think of this, it could-and did-serve in the "weighing" process to show
that plaintiff had not sustained the "burden of proof."
63
United States v. United States Gypsum Co., 67 F. Supp. 397 (D.D.C. 1946).
64Id. at 508-9. (Emphasis added.)
6 One hilarious example from the antitrust field I When erstwhile competitors are shown to
have been putting in identical bids, it is a fair inference that some sort of connivance has been
going on. But no, say defense counsel, since the bids are all exactly the same it really shows
that there was "perfect competition," and hence the case should be dismissed. And, strangely,
some courts have bought the argument. It would be more consistent with time-honored principles of fair trial, in such case, to ask the defendants to put in just a little evidence, under
oath, to support their suggestion. Otherwise, trial upon the facts becomes nothing but trial
by conjecture.
6United States v. United States Gypsum Co., supra note 63 at 509.
THE UNIVERSITY OF CHICAGO LAW REVIEW
[Vol. 27
Again, a subsidiary point in the Gypsum case was that the defendants had
conspired to eliminate jobbers in the distribution of gypsum board. There was
quite substantial evidence on the point, for USG had issued a price bulletin
specifying that thenceforth the price to jobbers was to be the same as that to
dealers. Moreover, among the subsequent exhibits was one in which a manufac-7
turer had wired that "Lee Story must be eliminated immediately as a jobber."
That would seem to make a prima fade showing, but the sender, an adverse
witness, testified on cross examination that his wire was not meant that way.
What he had wanted to say was that the manufacturer selling to Lee Story
should cease giving him "a jobber's discount," if that were the case, not that
Lee Story should be "eliminated."8 The court said this: "His [the sender's] explanation is consistent with his testimony that there was no elimination of job" Thus, once more, the
-.
bers but merely an elimination of their discount.
plaintiff had failed to sustain the burden of proof.
These illustrations may well appear unbelievable. But, in fact, they show
what can happen, and with an able court, when a trial judge lays aside his
judicial robes, midway through a trial, and undertakes to weigh plaintiff's evidence against what he imagines the defendant might prove. When the Gypsum
case finally reached the Supreme Court, counsel for USG had a difficult time
defending the trial court's findings. When he seriously urged that there was no
sufficient showing that jobbers had been eliminated from the industry, Chief
Justice Vinson leaned from the bench and said: "You mean, sir, that if the
jobbers wanted to, they were free to continue in business without a profit."
And that ended that.
The Supreme Court was also asked to decide whether, on motion to dismiss
pursuant to Rule 41(b), the trial court had properly ruled that it must weigh
the plaintiff's evidence and, further, whether the government must establish
its case by a preponderance of evidence. But Mr. Justice Reed found it unnecessary to do so: "We do not stop to consider those rulings. They are not of
importance in this case as we think the clear preponderance of evidence at the
conclusion of the government's case indicated a violation of the Sherman Act."70
ON CREDIBILITY
It may be that a trial judge engages in a "weighing" process of a different
order, though I doubt it, when he passes on the credibility of plaintiff's witnesses. Even so there is a question, not so much as to the weight to be given his
conclusions, although that, too, is in issue, but as to how and when he should
arrive at them. That is, should credibility be passed on at once--with all the
certitude of a woman's intuition-or should a court wait as impartially as may
67
69Ibid.
Id. at 493.
68Ibid.
States v. United States Gypsum Co., 333 U.S. 364, 388 (1947).
70 United
1959]
TBE PRIMA FACIE CASE IN NON-JURY TRIALS
be for the evidence to unfold? If experience had shown that trial judges are
quite infallible in such matters, the sooner they should "conclude" that a witness is trustworthy or not, the better. Much time could be saved that way.
In jury cases, of course, the trial court's part in determining credibility is
quite limited. Most jurisdictions treat credibility, like fact, as wholly a matter
for the jury to determine.7' But a considerable number give the court a screening power, comparable to that under the "substantial" evidence test. The formulations vary, but they come down to something like this: If the trial judge
decides that no reasonable person could properly rely on a witness' testimony
essential to plaintiff's case, he may direct a verdict. 72 Needless to say, the power
to direct a verdict, like the power to set aside a verdict 7 3 is one to be exercised
fairly, with a decent regard for the role of the jury. Trial by jury was instituted,
not because a jury is more experienced or intelligent than a trial judge, but, to
put it bluntly, as a necessary safeguard against biased or harsh judicial action.7 4
But in the non-jury case the trial judge is himself the trier of fact. And, since
he has the opportunity to see the witnesses and observe their "demeanor" on
the stand, great weight has always been given to his conclusions on credibility. 75
1See cases cited: 9 WiooRE, EVIDENCE § 2494 (3d ed., 1940); McBaine, Trial Practice,
31 CALi. L. REv. 454, 461 (1943).
72 For instance, in Gunning v. Cooley, 281 U.S. 90, 94 (1930), Mr. Justice Butler said:
"Issues that depend on the credibility of witnesses, and the effect or weight of evidence are
to be decided by the jury." But he first had said: "A mere scintilla of evidence is not enough
to require the submission of an issue to the jury. The decisions establish a more reasonable
rule 'that in every case, before the evidence is left to the jury, there is a preliminary question
for the judge, not whether there is literally no evidence, but whether there is any upon which
a jury can properly proceed to find a verdict for the party producing it, upon whom the onus
of proof is imposed.'"
7
3 It is usual to say that a verdict may be directed for a defendant when the trial court
would be bound to set aside a verdict for plaintiff on the ground that it was contrary to the
weight of the evidence. But, as Wigmore points out (9 WIGmORE, EVIDFNCE § 2494 (3d ed.
1940)), this test is neither helpful nor strictly accurate. Moreover, it is wholly inapt if sought
to be applied in ruling on a motion midway of a trial.
74 See discussion of the reasons for the Seventh Amendment, Clark & Stone, Review of
Findingsof Fact,4 U. Car. L. REv. 190 (1937).
"5Lord Shaw's statement in Clarke v. Edinburgh Tramways Co. (1919) H.L. 35, 36, is
often quoted: "When a judge hears and sees witnesses and makes a conclusion or inference
with regard to what is the weight or balance of their evidence, that judgment is entitled to
great respect, and that quite irrespective of whether the judge makes any observation with
regard to credibility or not. I can of course quite understand a Court of Appeal that says that
it will not interfere in a case in which the judge has announced as part of his judgment that he
believes one set of witnesses, having seen and heard them, and does not believe another. But
... in the ordinary case things are much more evenly divided; witnesses without any conscious
bias towards a conclusion may have in their demeanour, in their manner, in their hesitation,
in the nuance of their expressions, in even the turn of the eyelid, left an impression upon the
man who saw and heard them which can never be reproduced in the printed page." See also,
Judge Frank's remarks in Broadcast Music, Inc. v. Havana Madrid Restaurant Corp., 175
F.2d 77 (2d Cir. 1949).
THE UNIVERSITY OF CI-ICAGO LAW REVIEW
[Vol. 27
It would be nonsense, though, to suggest that trial judges are always right,7
or even always unbiased, in determining credibility. In the nature of things,
with the stress of argument in the conduct of a trial, with rulings to make on all
manner of points, with the court's own predilections about the nature of the
case, any trial judge, sitting alone, is very prone to take sides. Probably a jury
takes sides, too, though possibly for different reasons, but in its case every
precaution is taken to make sure that it will withhold judgment until the case
is finally submitted. Partly this is because a jury must reach a group decision,
but it is also because it would be unfair to do otherwise.
To get to the question, then, how does a judge determine credibility? The
demeanor of a witness, while of importance, is surely an overrated factor. 7 A
good witness may testify with poise and assurance to an untruth, or a half truth,
particularly if he knows there are no writings to controvert him;78 a poor witness may fumble and equivocate in telling the absolute truth. This is all commonplace. Essentially, it would seem, the process of determining credibility is
not much different from that of determining any other fact. That is, a witness'
credibility is really determined very largely by weighing his testimony against
other "facts" in the case: his own answers upon cross examination; the testimony of the witnesses for the other side; the writings which have been introduced as exhibits. Only rarely, if ever, may a court properly delve into its
uncontradicted witness' tes;timony by the soinner consciousness to test 7an
9
called inherent probabilities.
7 The trial judge in Quercia v. United States, 289 U.S. 466 (1933) called the jury's attention to the fact that defendant had frequently wiped his hands during his testimony: "It is
rather a curious thing, but that is almost always an indication of lying. Why it should be so
we don't know, but that is the fact." This was held to be reversible error. It would have been
equally so-though the fact might never have come to light-if the court had acted on that
basis when sitting without a jury. This and the other cases cited by Chief Justice Hughes in
the course of his opinion raise a perceptible doubt whether some appellate courts have not
overdone their deference to the trial judges' findings.
77 It ordinarily becomes critical only in the closely balanced case, after both sides have been
heard. In Powell & Wife v. Streatham Manor Nursing Home [1935] A.C. 243, 256, Lord Macmillan pointed out that the House of Lords must not hesitate to exercise its jurisdiction to
reverse when the 'rial court has erred on a question of fact, but then said: "Where, however,
as in the present instance, the question is one of credibility, where either story told in the
witness-box may be true, where the probabilities and possibilities are evenly balanced and
where the personal motives and interests of the parties cannot but affect their testimony, this
House has always been reluctant to differ from the judge who has seen and heard the witnesses, unless it can be clearly shown that he has fallen into error." See also, Flower v. Ebbw
Vale Steel Iron & Coal Co. [1936] A.C. 206, where the appeal was allowed.
78 Judge Hand's comments on this sort of thing in United States v. Corn Products Refining
Co., 234 Fed. 964, 978 (S.D.N.Y. 1916) are in point: "The documents were never intended to
meet the eyes of any one but the officers themselves, and were, as it were, cinematographic
photographs of their purposes at the time they were written. They have, therefore, the highest
validity as evidence of intention, and, although in many instances Bedford attempted to contradict them, his contradiction only served to affect the general credibility of his testimony."
71 See, Broadcast Music v. Havana Madrid Restaurant Corp., 175 F.2d 77 (2d Cir. 1949),
where the trial court's dismissal at the end of plaintiff's evidence was sustained.
1959]
THE PRIMA FACIE CASE IN NON-JURY TRIALS
In the Gypsum case80 Judge Stephens found all the adverse witnesses called
by the government to be quite credible, particularly in their answers upon cross
examination. But as to this, Mr. justice Reed said:
The government relied very largely on documentary exhibits and called as witnesses many of the authors of the documents.... On cross examination most of the
witnesses denied that they had acted in concert in securing patent licenses or that
they had agreed to do the things which in fact were done. Where such testimony is
in conflict with contemporaneous documents, we can give it little weight, particularly
when the crucial issues involve mixed questions of law and fact. Despite the opportunity of the trial court to appraise the credibility of the witnesses, we cannot, under
of this case, rule otherwise than that Finding 118 is clearly erthe circumstances
8
roneous. 1
In one view, the Supreme Court may be taken to have said that the Gypsum
witnesses, on cross examination, lacked credibility. That, of course, would have
cut the ground completely from under the trial court's elaborate analysis of the
evidence. But perhaps Mr. Justice Reed meant that, granting the witnesses'
credibility, their testimony did not square with the other, better established,
evidence in the case. That is to say, it simply lacked weight. For present purposes it is not necessary to decide which. It is surely evident that in either view
the trial court was engaged in substantially the same enterprise.
Ordinarily, by the time a plaintiff has rested, there can be little more than a
testing for consistency; the court's finding, then, at best, is one of prima facie
credibility. That is, "credibility," like any "fact," is but the end product of a
process of testing, and entitled to weight only according to the extent and accuracy of the testing. To put the matter very plainly, the findings of a trial court
upon "credibility," when made midway of a trial, are not entitled to the deference and respect traditionally accorded them when both sides have been
heard. Judges are not clairvoyant.12
SPEED AND HuRRY
It should be plain by now that the amended Rule 41(b) was not designed to
aid in getting at the truth of a controversy. If that were its purpose, the trial
court would have been admonished not to take sides early in a case; in fact, not
to "conclude" anything, either -as to "fact" or "credibility"--except in the
80 United States v. United States Gypsum Co., 67 F. Supp. 397 (D.D.C. 1946).
81United States v. United States Gypsum Co., 333 U.S. 364, 395-6 (1947). Finding 118
was to this effect: "The evidence... fails to establish that defendants associated themselves
in a plan to blanket the industry under patent licenses and stabilize prices, as charged by the
government."
82 See Boris,
TFxAL JUDGE 132 (1952): "I distrust my own surface reactions to a witness's
demeanor. He is seldom at ease; an adaptable knave will often accommodate himself more
readily to the rigid and unnatural climate of the courtroom than will an honest man. One
judge's impression of an open countenance may be another's impression of a shifty look. The
most depraved character I ever prosecuted had the beatific expression of a choirboy."
THE UNIVERSITY OF CHICAGO LAW REVIEW
[Vol. 27
most obvious cases-until the other side had been heard. Or, according to timehonored practice, until after plaintiff had put in his rebuttal, and defendant his
evidence in rejoinder. But this was not done. The new procedure, it seems, must
therefore be justified very largely, if at all, on the supposed need for speed, and,
I suppose, on a tender solicitude for the taxpayer's dollar.
The case of Global Commerce Corp. v. Clark-Babbit Industries,3 recently decided by Judge Hand, will illustrate the point. Here was a contract case which,
in the last analysis, turned on conversations between the two parties. Plaintiff
introduced the various letters and wires sent during the negotiations, from
which it appeared that by September 16th a "fairly complete proposed agreement"84 had been reached calling for shipment by defendant of 10,000 tons of
copra C.I.F. Mexican West Coast at a price of $110.25. Complications developed as to defendant's source of supply, as to the authority of defendant's
representative Myers, as to what, if any, performance bond should be required.
But, on September 25th, plaintiff wired saying offer "firmly accepted" and that
a letter of credit would be opened without performance bond. Defendant's
Vice-President Thomas, at this point, called plaintiff's representative Klein in
Mexico, discussed the deal, and invited Klein to New York.
Klein was called as a witness, and after the exhibits had been received and
he had testified to his conversations with Thomas, plaintiff rested. It would
seem that plaintiff had made out a prima facie case. judge Walsh, however,
dismissed from the bench, saying that plaintiff had not sustained "the burden
of proof." Perhaps he felt uneasy at holding defendant to a contract for so large
an amount without a written commitment on defendant's part, but, since
Mexican law was said to apply, 5 no writing was necessary.
It seems, though, that judge Walsh simply refused to accept Klein's version
of his talks with Thomas: "Now his testimony again doesn't sound in the words
of Thomas. He speaks in terms of legal conclusions, that Thomas promised to
deliver the 10,000 tons of copra without further delay as previously negotiated,
I think he said.""6 Judge Walsh did not find that Klein lacked credibility, though
he had doubts. What he said was this: "I can even see how over ten years of
thinking about a grievance that these conversations can almost fall into that
87
shape without deliberate distortion."
Again, speaking of Klein's testimony, judge Walsh said: "Now the fact is
that that sounds perhaps like a cable or a business letter. It does not sound like
83255
F.2d 105 (2d Cir. 1958). The case was heard by judge Walsh, without a jury.
8
4"Now as I view the transaction, between September 13th and September 18th, Myers
and Klein negotiated a fairly complete proposed agreement-I will talk about the copra firstcalling for delivery of 10,000 tons of copra against a letter of credit which in turn was to be
against a 5 per cent performance bond." Opinion of judge Walsh, unreported.
8 Global Commerce Corp. v. Clark-Babbitt Industries, 239 F,2d 716, 719 (1956),
8 Opinion of judge Walsh, unreported,
87 Ibid.
1959]
THE PRIMA FACIE CASE IN NON-JURY TRIALS
what Thomas said in a telephone conversation." 88s It is, of course, immaterial
whether Klein gave a literal version of Thomas' words; it would be enough if he
gave, as nearly as he could remember, their intent and meaning. But, the question is this, by what process-except that of sheer conjecture-could any trial
court possibly know what would "sound in the words of Thomas," or "like what
Thomas said," when Thomas had not been called as a witness?
There was, of course, much else in the case, but this was critical evidence,
and with all respect the court's disposal of it illustrates the grave danger in a
procedure which permits a trial court to "weigh" plaintiff's evidence, quite by
itself, to determine whether he has sustained the "burden of proof." In the
nature of things the trial judge must put his own doubts and conjectureswhich may or may not be well founded-into the balance against plaintiff's
proof. 89 And, to do that, comes perilously close to requiring plaintiff to establish
his case according to the rule of the criminal law, that is, beyond a reasonable
doubt. Surely the rule-makers did not intend-nor were they authorized to
make-so basic a change.
Nor, in this case, can it be said that much time was actually saved, except
that of the trial court. For, on appeal, the time of the circuit bench-Judges
Hand, Clark and Dimmock-was required to study the briefs, hear argument,
and write an opinion. It would have taken fewer judge-hours, certainly, and
contributed much to the fairness of the procedure, if the trial court had put the
defendant to his proof. Both Myers and Thomas were available as witnesses;
they had acted for the defendant; and, in perhaps a day, could have set matters
straight--one way or the other." But as to this Judge Hand said: "If the plaintiff had wished to secure any evidence from Thomas, Myers or any other of the
defendant's employees, it should have called them to the stand, in which event
it would not have been limited in its cross-examination-Rule 43(b)."'I
88 Commenting on Klein's testimony concerning an earlier conversation, Judge Walsh said:
"Now coming to his testimony as to this conversation, he testified that he told Thomas that
'We already sold the copra purchased in the call from Myers'. Now I don't believe he said
that. I mean that sounds to me like the retrospective thoughts of a lawyer-I mean of Klein,
whether he was a lawyer or not, or whether he is now or was at the time of his conversation,
-I think he still shows the effect of his early training .. " Ibid.
89 By way of further example: "I might say at the outset I don't understand Myer's activities. If he were a dishonest man I don't see what he had to gain by making proposals out
of sheer dishonesty if he didn't have a supply of copra or steel and yet offered the deal. The
only two other conclusions are either that he was insane or that he did have a supply which
was exhausted while Thomas and Klein haggled over the details of the transaction." With all
due respect, conjectures of this sort have no place in a trial upon the facts. There was nothing
in the record to show that Myers was dishonest, or insane, or that he did other than to promote a more or less usual business deal for the sale of copra. If judge Walsh entertained such
doubts, though, it would seem imperative that he should call the other side and get the truth.
"1The court in Alexander v. Rayson [1936] 1 K.B. 169, 178-9 points out the inconvenience
of stopping the case: "The witnesses... were presumably in Court and the expense of their
attendance had already been incurred." See text accompanying note 45 supra.
"1Global Commerce Corp. v. Clark-Babbitt Industries, 255 F.2d 105, 108 (1958). The
same point was made by judge Stephens in the Gypsum case, 67 F. Supp. 397,419 (D.D.C.
1946).
THE UNIVERSITY OF CHICAGO LAW REVIEW
[Vol. 27
Aside from being unrealistic, the suggestion is curiously inverted. The court
does not say so, but the implication is plain that, since a plaintiff may call the
defendant's witnesses and interrogate them as if upon cross-examination, it
follows that he must do so-or be charged with the trial court's adverse conjectures. Surely, that is going too far in the name of hurry.92 The purpose of Rule
43(b) was to give each party a fuller opportunity to bring out the facts of a
controversy to the end that cases might be decided on their merits, and less as
matter of chance.93 There was no thought, so far as I can discover, that the
parties should become "obliged to fish in their enemies' water."9 4 As a practical
matter, in fact, such a course would often be suicidal, particularly since the rule
permits the adverse party also to put leading questions to the witness, and, if
need be, to impeach him.99 It seems Rule 43 (b), like the original Rule 41 (b), was
drafted with the thought that you would pay some regard to the practice of the
last 150 years.96
Whether Judge Hand approved of the new procedure is not clear; he simply
said the third sentence of Rule 41(b) permitted the trial court to weigh plaintiff's evidence, without hearing the other side.97 However, he did refuse to fol"Nor is it apparent how any time is saved, for surely it takes about so long to produce a
given quantity of evidence, whichever party calls the witness.
"1See generally, Pike & Willis, The New FederalDeposition-DiscoveryProcedure,38 CoLrm.
L. REv. 1179 (1938).
11Judge Hand's expression, in a different situation: Schwartz v. United Merchants &
Manufacturers, Inc., 72 F.2d 256, 259 (2d Cir. 1934). (Emphasis added).
'5 FED. R. Civ. P. 43(b): "and the witness thus called may be contradicted and impeached
by or on behalf of the adverse party also, and may be cross-examined by the adverse
party.. . ." See generally, Johnson v. Baltimore & O.R.R., 208 F.2d 633 (3d Cir. 1953).
9"Many cases hold that when a plaintiff has produced substantial evidence to support his
charges, as plaintiff appears to have done in the Global Commerce case, a failure on the part of
the defendant as part of his case to call witnesses available to him and who knew the facts,
must weigh strongly against defendant. Mammoth Oil Co. v. United States, 275 U.S. 13, 52
(1927); Local 167 v. United States, 291 U.S. 293, 298 (1934); Interstate Circuit v. United
States, 306 U.S. 208 (1939). In Local 167 Mr. Justice Butler said: "The government introduced
substantial evidence which uncontradicted and unexplained tends to show that the conspiracy
and appellants' participation continued until the filing of the amended complaint. They were
present in court but failed to testify in their own defense. It justly may be inferred that they
were unable to show that they had abandoned the conspiracy.... Under the circumstances of
this case their silence rightly is to be deemed strong confirmation of the charges brought against
them."
" Judge Hand apparently thought Klein's testimony lacked credibility, though he had not
observed Klein's demeanor on the witness stand, and Judge Walsh, who had, did not say he
lacked credibility. It was just surmised that Judge Walsh was trying to be considerate: "we
are a little surprised at the moderation with which he expressed himself."
This compares with Judge Walsh's own polite conjectures as to Thomas' possible state of
mind when talking with Klein: "I think that Thomas wanted very sincerely to save Klein
any embarrassment from the transaction, and I think further that he would have been glad
to go ahead with Klein and Consumidores [Klein's buyer] if he could have gotten the supplies
to do it." Surely, this is all pure guesswork, and should have no place whatever in a trial upon
the facts.
1959]
THE PRIMA FACIE CASE IN NON-JURY TRIALS
low Benton v. Blair."'The trial court there, sitting without a jury, had also dismissed at the end of plaintiff's case, saying: "I am simply unable to accept as
true the plaintiff's version that on the first time he met the defendant Blair
in a two-hour conversation they made a contract, an oral contract, of the terms
and tenor that the plaintiff contends for.... ."" On appeal, Judge Tuttle said
the trial court was correct in its assertion that it must "weigh" plaintiff's evidence, but that "it was clearly error... to reject the uncontradicted, unimpeached and not inherently impossible or suspicious testimony of Benton .... "I
But it may be the New York bench thought Judge Tuttle had somehow lapsed
into the old practice, for he ended by saying that plaintiff's testimony made
out "a prima facie case as to the existence of a contract."10'
SPEEDIER SUMMARY JUDMENTS
One wonders from all this whether there has not been some waste of time in
the summary judgment procedure. That is, if a trial court can "conclude," from
a simple reading of plaintiff's affidavits, that a defendant may not be able to
make out a defense-regardless of whether an issue of fact is presented-why
should it not enter judgment for plaintiff at once? Surely it would be stultifying
to the court, and a waste of the taxpayer's dollar, to have to go on with a trial
when the court is not prepared to believe the defendant's evidence. Perhaps,
here too, the courts have simply failed to note that the rule--permitting the
other party to put in his proof-was probably developed to protect him in his
rights to a jury trial.
Some credence is given this idea by the comments of Mr. Justice Jackson in
Sarlorv. Arkansas NaturalGas Corp.1o2 There the question at issue was whether
the price of gas at the well head prior to 1930 "was above 3 per m.c.f.," whatever that means. The trial court had given summary judgment for the defendant, based upon his affidavits, but this was held to be error:
The Court of Appeals below heretofore has correctly noted that Rule 56 authorizes
summary judgment only where the moving party is entitled to judgment as a matter
of law, where it is quite clear what the truth is, that no genuine issue remains for
trial, and that the purpose of the rule is not to cut litigants off from their right of
trial by jury if they really have issues to try.03
Subsequent courts have followed the Sartor case, without regard to whether
a jury trial was involved.0 4 And, of course, Rule 56 itself has no special exception for those actions "tried by the court without a jury," as Rule 41(b) now
does. So, perhaps the rule-makers really overlooked a point. But, it is submitted, there is a more basic reason for denying summary judgment in these cases,
93228 F.2d 55 (5th Cir. 1956).
101Ibid.
91 Id. at 58.
101321 U.S. 620 (1944).
,10Id. at 61.
103 Id. at 627.
104 See, Dewey v. Clark, 180 F.2d 766 (D.C. Cir. 1950).
THE UNIVERSITY OF CHICAGO LAW REVIEW
[Vol. 27
as, indeed, for denying a motion to dismiss where plaintiff has made out a
prima facie case. That is, whether or not a jury is claimed, the very essence of
a fair trial is that there shall be a full hearing on disputed questions of fact. 5
Nor is this for the benefit of one adversary or the other; it rather is in the
public interest, to the end that controversies may be fairly decided on their
merits.
That "speed and hurry" are not reasons for limiting rights of this order has
nowhere been better stated than by Judge Hand, dissenting in CaliforniaApparel Creators v.Wieder of California.05 Saying, first, that an action for unfair
competition should be the last kind in which to invoke the remedy of summary
judgment, he went on:
Indeed, when I see, as I am constantly seeing more and more, the increasing disposition to make use of that remedy, I cannot help wondering whether there is not danger
that it may not rather impede, than advance, the administration of justice. It is an
easy way for a court with crowded dockets to dispose of them, and the habit of recourse to it readily becomes a denial of that thorough, though dilatory, examination
of the facts, on which justice depends even more than upon a studious examination of
the law; for a mistake of law can always be reviewed. Speed and hurry ought to be
the antipodes of judicial behavior. 10 7
SoME EQUITY PRECEDENTS
It is well now to re-examine the logic upon which the third sentence of Rule
41(b) was promoted. In essence it comes to this: Trial by jury is guaranteed
by the Constitution (in actions at law in the federal courts), in particular by the
Seventh Amendment, and hence it is for the jury, not the trial court, to determine questions of fact. And so, to quote Judge Stephens again, "if the reason
for the jury trial practice does not exist in non-jury trials, where the judge is
the trier of the facts, the jury trial practice ought not to be applied ...."I
Thus it followed, in Judge Stephens' view, that a trial court sitting without a
jury is bound to weigh a plaintiff's evidence, when ruling on a motion to dismiss,
in order to dispose of his case at the earliest opportunity.0 9 This, surely, is
logical nonsense. It assumes that the Seventh Amendment is the only reason
for applying the prima facie test." 0
If the assumption were correct the proponents of Rule 41(b), as amended,
05
The case of Porter v. Wilson, 239 U.S. 170 (1915), is difficult to handle; it seems to confound so completely the notion of a full trial on the facts with the supposed differences between
jury and non-jury trials.
106162 F.2d 893 (2d Cir. 1947).
'07 1d. at 903. See also, Arnstein v. Porter, 154 F.2d 464 (2d Cir. 1946).
108United States v. United States Gypsum Co., 67 F. Supp. 397, 417 (D.D.C. 1946).
109 Id.at 420.
110The English court, in Alexander v. Rayson, decided, quite as logically, that since a jury
would not weigh evidence midway in a trial, a judge sitting without a jury should not either.
See note 44 supra.
'
1o99
THE PRIMA FACIE CASE IN NON-JURY TRIALS
should be able to point to some equity precedent in its support. For the Chancellor, at least, has known from the start that he-and not a jury-is "the trier
of the facts." But, so far from finding support, it would appear there has been
no practice whatever in equity to "weigh" evidence midway in a trial. In fact,
it is black letter law that:
The case being set down for hearing on the bill, answer and proof, if defendant is
willing to risk his case on plaintiff's proof, or rather the failure of plaintiff to prove his
case, he should submit the case to the court for final hearing, and if he is not so satisfied, he should present what proof he desires or may be able to presentiu
The reasons for this practice are instructive: the more so since the Federal
Rules were drafted to afford a common procedure both for actions at law and
suits in equity."' In part, it no doubt came about from the nature of the early
equity proceeding, as one upon documentary evidence alone."' That is, with
both sides presented at the hearing there was little occasion for a defendant to
move to dismiss. But, probably, a more significant reason lay in the scope of the
appeal in equity, a matter to be discussed in detail in a moment. With a practice
to examine the whole record on appeal-instead of passing only on assigned errors--and with full freedom to set aside the trial court's findings," 4 there was
good reason to have both sides fully heard, to the end that the suit might be
disposed of once and for all by the appellate court."'
Perhaps the Illinois court, in Koebel v. Doyle,"' was making this point, when
it said: "To permit such a motion [to dismiss at the end of complainant's case]
would result in hearing a case by piece-meal, the sustaining of a motion resulting
in an appeal and on reversal another hearing on more evidence, followed, perhaps, by another appeal."" 7 But, it may be Justice Cartwright merely thought
WU
30 C.J.S. EQUITY § 579 (1942).
Clark, Two Decades of the FederalCivil Rules, 58 CoTum. L. Rav. 435 (1958).
112
"13
See, Chancellor Bland's opinion in Winder v. Diffenderffer, 2 Bland 166, 184-94 (Md.
1929).
114
See comment by Mr. Justice Jackson in District of Columbia v. Pace, 320 U.S. 698,
701-702 (1944).
"' The point was made by Etheridge, J., in Carter v. Studard, 118 Miss. 345,352,79 So. 225,
226 (1918), a case where the Chancellor had granted a motion for non-suit: "We have heretofore
condemned the practice... to sustain a motion at the conclusion of the complainant's evidence to strike out the evidence and give judgment for the defendant, without having the evidence of the defendant before the chancellor, and without requiring a defendant to elect to rest
his case upon complainant's evidence. This practice is wholly unknown in chancery practice
and procedure, and should not be indulged in the chancery court, as it undoubtedly protracts
litigation. If the entire evidence was in the record, we could frequently enter final judgment
here for one or the other of the parties, and this ought to be done whenever it is practical, to
the end that law suits should have an end."
u6 256 Ill. 610, 100 N.E. 154 (1912).
11 Id. at 614, 100 N.E. at 156. There is much in the court's language--this "cause was on
hearing before the chancellor for a final decision on the merits"-to suggest that the thrust of
the argument was against twice weighing evidence. If so, the equity objection to such motions
is really an objection to the amended Rule 41(b).
THE UNIVERSITY OF CHICAGO LAW REVIEW
[Vol. 27
it more expeditious to require the defendant either to rest or go on with his
18
case.
If the latter was his point, it was a dubious one, for surely some means should
be provided, even in equity, to discontinue a proceeding where-in the opinion
of the Chancellor-no person could reasonably say that the complainant had
made out a case. In modem equity practice," 9 much of the proof is presented
by witnesses in court-just as in a law action-and it is therefore difficult to see
why the same motion procedure should not be used in each case."2 At all events,
by the time Rule 41(b) was drafted, many states had reached this position,"'
and others-including Illinois2-changed their practice shortly afterward2 3
Thus, there is nothing in chancery practice to support the point that a trial
judge, sitting without a jury, must "weigh" evidence when ruling on a motion
to dismiss. Indeed, one senses that-all else aside-the early Chancellor, presiding as he did over a court of conscience, had certain scruples at engaging in
such an enterprise, when the other side had not been heard, or had not closed its
case." 4 But a motion to test whether a complainant has introduced "substantial evidence" to support the necessary allegations of his bill is something else
again. That far the draftsmen of Rule 41(b) were entitled to go, and that far
only, without doing violence to the spirit of the prior equity practice.
Some states, it is true, have now gone even further. In its 1955 Civil Practice
118
Many defendants nevertheless persisted in moving to dismiss, or to make some other
motion, as to strike evidence. If the Chancellor's action in dismissing the bill was sustained on
appeal, there was no harm done. Otherwise, the apparent harshness to a defendant with a
good defense was sometimes softened by remanding the case with instructions to take further
testimony. See Sundlun v. Volpe, 62 R.I. 55, 2 A.2d 875 (1938), and cases discussed.
11 Equity R. 46, 226 U.S. 657 (1912), which require chancery courts to receive testimony
in open court, was announced in 1912. See generally, Clark & Stone, Reviev of Findingsof
Fact (1937) 4 U. Car. L. Rv. 190, 203-04.
,21In fact, even as early as 1885, a lower Illinois court had approved a motion to strike
plaintiff's evidence: "It is well settled in suits at law that a motion to strike out all the plaintiff's evidence is in the nature of demurrer to the evidence, and like such demurrer admits
not only all the facts proved but also every conclusion which the jury might fairly and reasonably have drawn therefrom. We see no reason why the same rule should not apply to a suit in
chancery, especially one like this, where the witnesses are all examined in open court...
Heiderich v. Heiderich, 18 Ill. App. 142, 143 (1885).
121 The Federal court in Cook v. Klonos, 164 Fed. 529, 535 (9th Cir. 1908), took a similar
position: "In an action at law.., defendant, for the purpose of the motion, admits... facts
which may naturally and rationally be inferred from the facts proven. There is no reason why
the rule should not be applicable to a like motion to dismiss in an action of an equitable
nature."
122 See Johnson v. Johnson, 313 Ill. App. 193, 39 N.E.2d 389 (1942). The amended Rule
64, sub. 4, of the Civil Practice Act, made it clear that a defendant in equity, at the close of
plaintiff's case, might "move for a finding in his favor" or "to dismiss the suit for want of
equity," without being deemed to have rested. ILL. REv. STAT. Ch. 110, § 188(4) (1949).
123 See Phillips v. Phillips, 215 Md. 28, 135 A.2d 849 (1957), and cases cited.
224A chancery court traditionally was interested in the fairness of its determinations, not
in speed and hurry. See remarks of Judge Smith quoted in text accompanying note 130 infra.
1959]
THE PRIMA FACIE CASE IN NON-JURY TRIALS
Act, Illinois bluntly provides, though upon what course of reasoning does not
appear, that a judge sitting without a jury, whether at law or in equity, "shall
weigh the evidence ' us when ruling upon a motion to dismiss at the close of
plaintiff's case. Probably the rulemakers were caught in the spell of the arguments advanced for the amended Rule 41 (b).us But they went the Federal rulemakers one better, for the trial court was not given an option, either to dismiss
or to defer ruling until the other side had been heard. He "shall weigh the evidence."12 1
Arkansas, which has also changed its chancery practice, now allows a defendant-without being required to rest-to file a motion, at the close of the
plaintiff's case, challenging the sufficiency of the evidence to warrant the relief
prayed. Nothing was said, though, as to how the Chancellor should evaluate
the plaintiff's evidence. But in Werbe v. Holt,n8 where that was a point in issue," 9
Judge George Rose Smith refused to approve a weighing procedure:
The American courts have always followed the theory of an adversary trial. In such
a trial the parties are placed on equal terms and each develops his own proof by his
own witnesses, though of course the party having the burden of proof must establish
a printafaciecase before his opponent need go forward with the evidence. The minority
conception of a demurrer to the evidence is contrary to the traditional procedure in
adversary trials, since the defendant is given an advantage. He has the opportunity
of twice submitting the case to the trier of facts.... Furthermore, in many instances
the plaintiff's printafaciecase must necessarily be somewhat weak, for the reason that
only the defendant himself may be able to supply details needed to complete the picture. If the case goes to the trier of the facts on the plaintiff's proof alone, the defendant
has the advantage of not exposing weaknesses in his own armor unless called to the
witness stand by his adversary. For these reasons we have no hesitancy in adopting
the majority rule as to the function of a demurrer to the evidence. 130
11z
Rule 64 (5), Civil Practice Act.
16InPhillips v. Phillips, 215 Md. 28, 135 A.2d 849, concurring opinion at 136 A.2d 862
(1957), Chief Judge Brune, likewise, urged the rulemakers to adopt the new Federal procedure
to give the Chancellor power to "weigh" the plaintiff's evidence midway in a trial. But the
reason he urged for the change was rather an argument for adopting the "substantial evidence"
test: "The argument that because a court of equity is a court of conscience, the Chancellor
should hear both sides before deciding the case is of little force or applicability in a situation
of this sort, as I see the matter. It would, I think, be a strange manifestation of conscience to
require a defendant to meet and disprove a case which the Chancellor, after a full hearing of
the plaintiff's side of the matter, considered to be groundlessor illfounded." (Emphasis added.)
127Supra note 125.
-8 217 Ark. 198, 229 S.W.2d 225 (1950).
129 Id. at 200, 229 S.W.2d at 226. The court stated the precise issue: "In the case at bar
this question is of primary importance, for the appellant's proof was undoubtedly sufficient
to raise a jury question had the case been tried in a circuit court. But if the problem is where
the preponderance lay, a much closer question is presented."
130
Id. at 202-03, 229 S.W.2d at 227. Perhaps Judge Smith was influenced by the fact that
defendant's motion was in the nature of a demurrer. But, whether so or not, his comment
would apply equally to a motion to dismiss, which traditionally was treated much like a
demurrer. See note 2 supra.
THE UNIVERSITY OF CHICAGO LAW REVIEW
Tim
[Vol. 27
JuRY's MANTLE
It was much controverted, when the Federal Rules were drafted, whether
the law or the chancery practice should apply on appeal.131 One thing was
agreed, the procedure should be the same in each case, if that were at all possible. After full debate the chancery practice was adopted.13 2 Not that there
should be a complete trial de novo,133 as in early chancery practice, but definitely to adopt the modern equity procedure, under which the reviewing court
is free to examine the complete record on appeal and-more important for
court made an unwise,
present purposes-to reverse if, in its opinion, the trial
13 4
i.e. wrong, decision, either on the law or on the facts.
The language adopting this position is found in Rule 52(a), which applies to
"all actions tried upon the facts without a jury... ."I" After providing that the
trial court "shall find the facts specially. . . ," the rule goes on to admonish
reviewing courts that: "Findings of fact shall not be set aside unless clearly
erroneous, and due regard shall be given to the opportunity of the trial court
to judge of the credibility of the witnesses." ' , Why the rule was not phrased in
the affirmative, that is, that a reviewing court shall set aside findings which are
clearly erroneous, does not appear, but probably the actual meaning would be
the same if it were stated that way. The reviewing court, thus, is given full
power-indeed, is bound' 37 -to over-turn findings of fact which are, in its
opinion, "clearly erroneous."
131See, Clark & Stone, Review of Findings of Fact, 4 U. CHi. L. Ryv. 190 (1937).
13
MooRE, FEDERAL RuLEs 246 (1956).
131See, United States v. du Pont & Co., 351 U.S. 377, 381 (1956), where Mr. Justice Reed
said: "We do not try the facts of cases de novo."
1 4 The Supreme Court has made it clear in many cases that it will not ordinarily engage
in an extended review of the record. United States v. Yellow Cab Co., 338 U.S. 338 (1949);
United States v. Timken Roller Bearing Co., 341 U.S. 593 (1951). But this cannot be other
than a self-imposed restriction, perhaps dictated by the pressure of Supreme Court business,
-Court of Appeals business is something else again-for there is nothing in Rule 52(a) or
the prior equity practice to require it. Besides, it has no bearing on how the court should review
such part of the record as it does examine.
The so-called Big case presents special problems, but, even so, it should not dictate general
procedure. See McAllister, The Big Case, 64 Hav. L. REv. 27 (1950); Seminar on Protracted
Cases, 21 F.R.D. 401 (1957). See also Gottesman v. General Motors Corp., 268 F.2d 194 (1959),
construingthe Interlocutory Appeals Statute (28 U.S.C.A. § 1292(b)).
116Ibid. (Emphasis added.)
115Fed. R. Civ. P. 52(a).
137 Lord Macmillan stated the English practice as follows: "If the judgment given in plaintiff's favour in the Court of first instance had been based on the verdict of a jury your Lordships' task would have been relatively easy, for there was plainly ample evidence on which a
reasonable jury could have found for the plaintiffs and the verdict must have stood. But the
case was tried by a judge sitting alone, and on appeal from the decision of a judge the Court
of Appeal and this Court have a duty to exercise their jurisdiction as tribunals of appeal on
fact as well as on law, a jurisdiction which your Lordships have never hesitated to exercise
when satisfied that the Courts below have erred on a question of fact." Powell and Wife v.
Streatham Manor Nursing Home [19351 A.C. 243, 256. See also, the court's comment in
Montgomerie & Co. v. Wallace-James, [1904] A.C. 73, 75.
19591
THE PRIMA FACIE CASE IN NON-JURY TRIALS
In concept, Rule 52(a) represented a sharp departure from the prevailing
appellate procedure on the law side. There, for many years-without warrant
of any Constitutional mandate-the mantle of the jury had been put upon the
not unwilling shoulders of the trial judge. That is, the findings of the trial
judge, as a "trier of the facts," were given the same finality as the verdict of a
jury, 13 and could not be set aside except where an appellate court could find no
"substantial evidence" in their support. It was irrelevant that, in the opinion
of the appellate court, the facts warranted an opposite result. The new rule,
though, lodges authority with the appellate court; its opinion, and not that of
the trial court, is to prevail, except only insofar as a "due regard" for the
opportunity of the trial court to judge of "credibility," and a proper respect
for the trial court's opinion, should stay its hand.
Notwithstanding the rule-maker's purpose, the words, "clearly erroneous,"
have had a curiously inhibiting effect.139 They seem to say, to the uninitiated
at least, that appellate court powers have been limited, not extended; the trial
court may no longer be reversed on the facts, it seems, except its findings are
clearly erroneous. And so, some courts are back to the old "substantial evidence" test; if anything, more rigorously applied than ever. For how in common
sense, it is asked, may the trial court's findings be dearly wrong, if there is to
be found some fact, or some inference therefrom, which tends to support them?
To illustrate, the Seventh Circuit, in Penn-Texas Corp. v. Morse,140 recently
reiterated its settled position that Rule 52(a) when read with 41(b) adopts the
"substantial evidence" test. Here was an equity case brought to enjoin the
issuance of roughly 169,500 shares of stock. The trial court dismissed at the
close of plaintiff's evidence and, on appeal, Judge Finnegan sustained its action,
pointing out that the trial court "had the primary function of finding the facts
and choosing from among conflicting factual inferences those which he con4
sidered most reasonable."' '
That much, of course, is true, but the court then went on to say:
Under such circumstances our power is limited to a determination of whether those
inferences and conclusions have any substantial basis in the evidence. If such a basis
is present the process of judicial review is at an end... and the finding of the District
Court must be accepted by this court.... And even where there is no dispute about
the facts, if different reasonable inferences may fairly be drawn from the evidence, we
I's
"Where a case is tried by the court, a jury having been waived, its findings upon questions of fact are conclusive in the courts of review, it matters not how convincing the argument that upon the evidence the findings should have been different." Per Shiras, J., in
Dooley v. Pease, 180 U.S. 126, 131 (1900). See also United States v. Jefferson Electric Co.,
291 U.S. 386, 407 (1934).
139
For an extremely lucid explanation of the difference between the two tests, see Stern,
Revkw of Findings,58 HAPv. L. REv. 70, 80-81 (1944).
140 242 F.2d 243 (7th Cir. 1957).
41
d.at 247.
THE UNIVERSITY OF CHICAGO LAW REVIEW
[Vol. 27
are forbidden to disturb the findings based on such inferences unless they are clearly
erroneous.4
Of course this shows a profound disregard of the background and purpose
of Rule 52(a).14 1 Whether there is substantial evidence to support the trial
court is beside the point. Moreover, this was made quite clear in the Gypsum
case, 44 when Mr. Justice Reed said:
The practice in equity prior to the present Rules of Civil Procedure was that the findings of the trial court, when dependent upon oral testimony where the candor and
credibility of the witnesses would best be judged, had great weight with the appellate
court. The findings were never conclusive however. A finding is "clearly erroneous"
when although there is evidence to support it, the reviewing court on the entire evidence
is left with the definite and firm conviction that a mistake has been committed. 4l
Rule 52(a) was drafted with no th6ught that it would apply to the findings
of a trial court made upon purportedly "weighing" evidence midway in a trial.
It- stated a chancery practice, and there had been no practice of that sort in
chancery. Moreover, since the rule was designed to deny to a trial court's findings the degree of finality which has always been accorded to a jury's verdict
after both sides have been heard, it would seem, a fortiori, that findings made
midway in a trial should have no greater sanctity. Or, to state the point another way, even though a trial court when sitting without a jury is "a trier of
the facts," it is not privileged then, or at any stage of the case, to wear
the jury's mantle.
DEFERENCE To APPELLATE. CouRTs
One wonders just how the proponents of amended Rule 41(b) expected to
square it with Rule 52(a). Suppose a trial court does "weigh" evidence midway
in a trial, and make findings of fact, what does that accomplish? Very little,
indeed, except to cause delay and expense, if such findings may be set aside
as "clearly erroneous," even though supported by "substantial evidence." That
is, this is true, unless Rule 41(b) is to be construed as having amended Rule
52(a) to read to this effect: "Findings of fact shall not be set aside unless clearly
erroneous, and where a trial court has purported to weigh evidence at the dose of
42
Id. at 247. This language was taken from judge Kerner's opinion in Gaytime Frock Co.
v. Liberty Mut. Ins. Co., 148 F.2d 694, 696 (1945). Judge Kerner, in turn, took it, in the
main, from the opinion of Mr. justice Murphy in Commissioner v. Scottish American Investment Co., 323 U.S. 119, 123-124 (1944). But that case had to do with an appeal from the Tax
Court. And, what the Seventh Circuit failed to note, is that Congress provided specially many
years ago that findings of the Tax Court-presumably because of its expertness in the fieldmay not be overturned when supported by "substantial evidence." There is no statute giving
a similar measure of finality to findings of a trial judge in civil cases generally.
143For a good discussion, see ILsEN & HONE, FederalAppellate Practice,in FED. RULEs
or CIvir PROCEDuRE, 359, 418 et seq. (Rev. ed. 1947).
144333 U.S. 364 (1948).
'
145 Id. at 395. (Emphasis added.)
19591
TBE PRIMA FACIE CASE IN NON-JURY TRIALS
plaintiffs case, they may not be set aside at all, if there is any substantial evidence
in their support. .. " But Rule 52(a) was not amended in that way, and the
incongruity of making such an exception in favor of mid-trial findings should
at least give a court pause before improvising such an amendment.
More likely the proponents were trading on a certain lack of understanding
-plus a little inertia-on the part of the appellate courts. Perhaps the rulemakers numbered some who were still not persuaded to the wisdom of extending
appellate review, as is provided by Rule 52(a). If so, here was a way to bring
back the old limited review which had prevailed at law-and even to make it
apply in equity cases-all under the guise of adopting a liberal rule to speed up
trial. If such was their purpose, they succeeded better than they should have
done. But why stop part way; if the trial court's findings at mid-trial are entitled to the finality of a jury's verdict, why not abandon entirely the notion
of a full appeal as in chancery practice and give a similar finality to the court's
findings at the end? That, too, would simplify the work of the appellate courts,
and so promote an earlier disposition of cases.
It is not proposed to examine in more than brief detail the reasons for the
"clearly erroneous" test stated in Rule 52(a). Basically, it rests on the principle
that in our jurisprudence no litigant, who has more than a semblance of a case,
is to be thrown out of court by the ipsi dixit of a single judge. Congress has made
generous provision for a further hearing. Nor can it be said that an appellate
bench is less competent than the trial court to pass on the law and the facts of a
case. Indeed, it is usually more able to do so, for the appellate bench is relieved of
the hurry of daily trial work; there is the benefit of consultation; and, more
important, the case may be seen in fuller perspective than often is possible to
a single trial judge. Only to the extent that the trial judge's findings depend on
matters of credibility-after both sides have been heard-should a "due regard"
for his opportunity to see and hear the witnesses restrict review.
The verdict of a jury, on the other hand, is a very different thing. The essence
of the right to trial by jury lies in the assurance that a litigant shall have the
judgment of a panel of his peers upon disputed questions of fact.'46 Likewise in
the case of many administrative agencies, though for different but similar
reasons, Congress has provided that findings may not be set aside, except where
there is no "substantial evidence" in their support.147 In these areas it is not
open to an appellate court to say that we view the facts differently. 48 But a
146Galloway v. United States, 319 U.S. 372 (1943), "Since judicial review of findings of
trial courts does not have the statutory or constitutional limitations on judicial review of
findings by administrative agencies or by a jury, this Court may reverse findings of fact by
a trial court where 'clearly erroneous'." Per, Mr. Justice Reed in United States v. United
States Gypsum Co., 333 U.S. 364, 395 (1948).
14 See, NLRB v. Columbian Enameling & Stamping Co., 306 U.S. 292, 300 (1939), for discussion of the similarities between the rule for juries and that for administrative agencies. See
also Brown, Fact and Law in JudicialReview, 56 HARv. L. REv. 899 (1943).
'sStem, Review of Findings of Administrators, Jidges and Juries, 58 HARV. L. REv. 70,
73-78 (1944).
THE UNIVERSITY OF CHICAGO LAW REVIEW
[Vol. 27
single trial judge enjoys no unique status; he can point to no Constitutional
provision to give special weight to his findings; he has no Congressional sanction
giving finality to his actions. Nor is he appointed because of any special competence; his training and background are those of the appellate judge, no more
and no less.
Thus, not only is the way open, but it would seem to be in the public interest
to make sure that an appellate court does have freedom to review-and if
need be to set aside' 4 -the determinations of a single trial judge. After all, the
appellate court has a national (or state-wide) jurisdiction; it has the prime responsibility of making sure that the laws of the realm are enforced throughout
its length and breadth, equally and fairly. That a given policy, judge-made or
statutory, should result in one decision in one trial court and on much the same
set of facts in an exactly opposite one in another, is surely not tolerable. Yet,
under the old "substantial evidence" rule,5 0 if each trial court could point to
some evidence which might impress a reasonable man-as each well might do
in any close case-an appellate court could do very little to reconcile the two
decisions. That surely tended to hodge podge, not justice according to law.
The "clearly erroneous" test, as pointed out above, was not adopted casually;
it was carefully chosen to give the appellate courts power in law cases to take
greater responsibility than theretofore. 5' And so, by similar token, the amendment to Rule 41(b) can only be regarded as a step in the wrong direction. It
operates to set the judgment of the trial court above that of the appellate court,
whereas Rule 52(a) was designed to accomplish the exact opposite. When a trial
judge passes on a motion to dismiss at the end of a plaintiff's case, therefore, a
due deference to the appellate court 5' should impel him to ask: Is this evidence,
when viewed in a light most favorable to plaintiff, sufficient that an appellate
court might reasonably decide in plaintiff's favor? If so, the motion should be
denied.
This, of course, is essentially the test used in jury cases. It would serve the
same basic purpose, that is, to insure that the appellate court, like the jury,
should have the evidence upon which to reach its own decisions. That the trial
"I This, of course, was the practice in chancery, and some appellate courts recognize, to a
degree, that they now have the same responsibility in hearing appeals at law, when the case
is tried by a court without a jury. See Judge Frank's analysis in Orvis v. Higgins, 180 F.2d
537, 539 (2d Cir. 1950). Cf. comment by Clark, Special Problens in Drafting and Interpreting
ProceduralCodes and Rules, 3 VAInD. L. REv. 493, 506 (1950).
150The "substantial evidence" rule means different things at different times, but the test
does not change. When used in passing on a motion to dismiss, it tends to put the defendant
to his proof; when used after a verdict, it tends to give finality to the decision. What the proponents of amended Rule 41(b) seem to have wanted was to throw out the first application,
but to appropriate the second. They were wrong on both counts.
1', For the arguments, pro and con, see Clark & Stone, Review of Findingsof Fact,4 U. CHI.
L. REv. 190 (1937).
12 Deference is a quality which should work both ways; it is not all owed to the trial court.
Rule 41(b) gives the trial court this option.
19591
THE PRIMA FACIE CASE IN NON-JURY TRIALS
125
judge may have "concluded," after hearing only one side, that he would decide
for the defendant, 153 is interesting, but not too important. If he knows that
there is substantial evidence in the record upon which an appellate court might
well decide otherwise, it is a plain perversion of fair trial to purport to "weigh"
the plaintiff's evidence, find it wanting, and then to grant defendant's motion
to dismiss. In such case, as when trial is had before a jury, it is the first function of the trial judge to hear all of the facts, a difficult and most important
task-but one "on which," we are assured, "justice depends even more than
upon a studious examination of the law; for a mistake of law can always be re54
viewed."'
FINALLY
There is a certain rhythm in legal matters, as in most social phenomena. Perhaps the day is not yet, but it does seem that the heresy set afoot by Hoyt,
C. J., in 189615 5-that a trial court is to be encouraged to "weigh" evidence
midway in a trial-should have about run its course. No single argument advanced for the proposal, it seems, will bear careful analysis. 55 And, the unstated
central idea, that the doubts and conjectures of a trial judge may properly be
put in the balance and weighed against the plaintiff's facts, established by ex15 7
hibits and sworn testimony, is completely foreign to our notions of fair trial.
The obvious course is to delete the third sentence of Rule 41(b), at the earliest
opportunity.15 A misbegotten offspring of an unseemly desire for speed and
hurry, it has no place in our system of justice. It strikes at the citizen with a
substantial case; the one who least of all should be summarily put out of court.
There are many ways of expediting the trial of cases, 59 without bartering away
153An impulsive judge, no less than a "timid" or a "biased judge"-if I may add a psychiatric category to ir. Justice Frankfurter's dictum-"is intrinsically a lawless judge." Wilkerson v. McCarthy, 336 U.S. 53, 65 (1949).
114
Supra note 107.
15Ibi.
166
Of course, as Arnold has said, argumentative attack on such a device as Rule 41(b),
related as it is to deeply rooted notions of limited review, may well be pointless; "if the emotional need for the institution is strong," argument "does not change the institution; instead
it complicates its jurisprudence or its theology, until the attack is deflected and loses its force
in a maze of learning." Arnold, Apologiafor Jurisprudence,44 YALE L.J. 729, 748 (1935).
I5" Nor would it help matters to shift to a "probability" or "degree of belief" test. See, however, McBaine, Burden of Proof: Degrees of Belief, 32 CALIF. L. Rlv. 242 (1944). "Probability,"
or a "belief" founded on probabilities, has no place midway in a trial upon the facts, that is,
however strange or improbable a plaintiff's case may seem to be, it also may be based on cold
fact. Truth, it is said, is often stranger than fiction. It is no more respectable, therefore, to
base a "belief" on conjectures as to what a defendant probably might prove, than it is to
"weigh" such conjectures against plaintiff's evidence.
155 It would be a wise course to replace the third sentence with the following: "In ruling
upon the defendant's motion the court shall view the evidence and all inferences to be reasonably drawn therefrom in a light most favorable to the plaintiff." See text at note 12 supra.
1'9
See particularly the recent study by ZEISEL, KALVEN & BucEraOLz, DELAY iNrTn
COURT (1959).
126
THE UNIVERSITY OF CHICAGO LAW REVIEW
the rights of litigants to have a full hearing of their controversy. And, incidentally, it is not for the courts--or the rulemakers-to engage in a bartering enterprise; Congress has control of the purse strings and, so far as I know, has never
been niggardly when an issue of fair trial was before it.
It was not the purpose of this study to examine how the third sentence of
Rule 52(a) has worked in practice. But it is apparent that it too has worked
badly. It seems, therefore, that the rule-makers, when next they convene,
should seriously consider making the rule say precisely what it was intended
to mean. The following is suggested:
Findings of fact shall be set aside when clearly erroneous, and shall be regarded as
being clearly erroneous, when, although there may be substantial evidence in their
support, the reviewing court, on examining the entire evidence before it, and after
giving due regard to the opportunity of the trial court to judge of the credibility of
the witnesses, is left with the definite and firm conviction that a mistake has been
committed 6 0
And so, it would seem the traditional mid-point test of whether a prima fade
case has been made out, instead of being a "refinement of technicality," is a
necessary safeguard of justice, as much in court cases, as in those tried before
a jury.
160This definition of "dearly erroneous" is that stated by Mr. Justice Reed in United States
v. Gypsum Co. 333 U.S. 364,395 (1947); supra note 145. But, where the Justice says a finding
may be set aside "although there is evidence to support it," the word "substantial" has been
added. This makes the test explicit and, in the writer's view, does not change its meaning.
Also the words "before it" have been added to the phrase on "the entire evidence." This, too,
probably does not change the Court's meaning; at the same time it recognizes that the Court
has a free hand to review all or only part of a record on appeal as justice may require, or as
the Court's time will permit.